From: Junio C Hamano <gitster@pobox.com>
Date: Mon, 01 Sep 2008 19:33:48 GMT
Subject: Re: What's cooking in git.git (Aug 2008, #09; Sun, 31)
Message-ID: <7vprnnvexv.fsf@gitster.siamese.dyndns.org>
In-Reply-To: <fcaeb9bf0809010116p68a638dck4f8e31d49e51c3fb@mail.gmail.com>

"Nguyen Thai Ngoc Duy" <pclouds@gmail.com> writes:

> On 9/1/08, Junio C Hamano <gitster@pobox.com> wrote:
>>  * nd/worktree (Sat Aug 30 16:15:32 2008 +0700) 5 commits
>>   - setup_git_directory(): fix move to worktree toplevel directory
>>   - update-index: fix worktree setup
>>   + read-tree: setup worktree if merge is required
>>   + grep: fix worktree setup
>>   + diff*: fix worktree setup
>>
>>  The last two fail test suite.
>
> "make test" on pu did not fail on my machine. What tests failed?

*** t1006-read-tree-m-i.sh ***
* FAIL 1: setup
	
	
		mkdir it.git &&
		(
			GIT_DIR=. &&
			export GIT_DIR &&
			cd it.git &&
			git init &&
			M="1 2 3 4 5 6 7 8" &&
			for i in $M 9; do echo $i; done >file &&
			cat file >elif &&
			git update-index --add file elif &&
		...

which is part of my private test suite.

While it is a change in behaviour, after looking at what the test does
again, I do not think this use case is worth supporting.